 |
Pioneer research
blazes the trail
While defining new boundaries at the edges of
discovery, engineering faculty guide their students to the frontier
by Sara Peters
The mission statement of the University sets challenging
goals. Researchers at Princeton must not forget the legacy of
excellence they are expected to uphold as frontiersmen of discovery,
servants of the global community, and educators of the next
generation.
It is clear that those at the School of Engineering and Applied
Science (SEAS) have taken these missions to heart. SEAS research
is blooming. SEAS expenditures for sponsored research equaled
$39 million in 2001, increasing 15 percent over 2000 expenditures.
The number of patent applications stemming from SEAS research
increased by 50 percent in 2001.
These numbers include research done within the six SEAS departments,
plus collaborative work by faculty in the Center for Photonics
and Optoelectronic Materials (POEM), the Princeton Materials
Institute (PMI), the Princeton Environmental Institute (PEI),
and Princeton Applied and Computational Mathematics (PACM).
One hundred full-time professional research and technical staff
facilitate this explosion in research.
With money flooding in and research efforts swelling, SEAS faculty
and staff have made special efforts to keep the University mission
from being drowned in the surge.
Discovery
| “Princeton
University strives to be both one of the leading research
universities and the most outstanding undergraduate
college in the world. As a research university, it seeks
to achieve the highest level of distinction in the discovery
and transmission of knowledge and understanding. At
the same time, Princeton aims to be distinctive among
research universities in its commitment to undergraduate
teaching. Through the scholarship, research, and teaching
of its faculty and the many contributions to society
of its alumni, Princeton seeks to fulfill its informal
motto: ‘Princeton in the nation’s service
and in the service of all nations.’”
|
Many scientists retain the same winsome curiosity that caused
them to ask “Why is the sky blue?” and “How
do birds fly?” when they were children. These people
conduct research to fulfill a yen to answer some of mankind’s
most ancient questions.
Professors N. Jeremy Kasdin ’85 and Michael Littman
from the Department of Mechanical and Aerospace Engineering,
and Robert Vanderbei from the Department of Operations Research
and Financial Engineering (ORFE) are part of a research effort
that may help discover answers to a question asked throughout
the ages: Are we alone in the universe?
The research group is working to develop a telescope that
will find and identify planets that are similar enough to
Earth to support life as we know it. The group is one of four
teams competing to plan and execute the mission for the National
Aeronautic and Space Administration (NASA), which plans a
2012 launch.
Teaching
The ultimate goal of a university is to prepare the leaders
of tomorrow, and researchers must remember this. At SEAS,
special efforts are made to bring the lab into the classroom,
and the classroom into the lab.
ORFE Professor Warren Powell ’77, brings his research
into the classroom by requiring students to use advanced tools
developed in his lab to simulate operation of an efficient
orange juice business.
The success of the “Orange Juice Game” in this
class is an example of how innovative research can lead to
a richer educational experience for students.
Service
Since the time of Woodrow Wilson, Princeton University’s
unofficial motto “Princeton in the nation’s service
and in the service of all nations” has pervaded the
character of the campus, and has been reflected over the globe
with improvements to our daily lives.
Professors in the Department of Civil and Environmental Engineering
are seeking better information about global climate change:
How and when it will affect life on Earth.
Professor Eric Wood’s research group is helping create
better climate models to paint a clearer picture of this global
concern.
Professor Ruby Lee in the Department of Electrical Engineering
intends to make our lives easier by improving computers—and
she wants to start from square one.
She aims to build mechanisms for handling security and new
media directly into the computer architecture, rather than
heaping on more software.
This 11-page series highlights some of the researchers and
educators in the SEAS, who are working daily to further the
purposes of discovery, education, and service.
Crucial
cogs not always easily spotted
Professional research, technical staff key to success
Observe your
clock. See the dial, the numbers, and the hands. Hear the
ticking and the tocking and the snap of the minute hand as
it locks into place. Move closer, and you may hear a soft
whirring. Although you cannot see them, that soft whirring
is the sound of many wheels and cogs inside the clock, which
are hard at work to keep things running.
Now think of the School of Engineering and Applied Science
(SEAS) as a big clock. Perhaps a sturdy, elegant grandfather
clock. Inside there are 100 of these wheels and cogs, tucked
away in laboratories behind the unassuming term Professional
Research and Technical Staff (PRTS).
The PRTS work every day to make research happen and help maintain
Princeton University’s position as a top research school.
Generally they work for a specific professor and add their
skills and knowledge to the intellectual pot by performing
experiments specified by the professor’s goals.
In addition to these 100 full-time researchers, there are
many visiting research and technical staff on the payroll.
These researchers are a diversified group, coming from many
countries, professional backgrounds, and disciplines.
For example, Elmer Ledesma, a research staff member, translates
classical literature from the original Latin in his spare
time. Ben Shedd, a technical staff member in the Department
of Computer Science, is an Academy Award-winning short-filmmaker.
Their work varies widely from lab to lab. The job description
of a PRTS member may or may not include: teaching; lab setup,
maintenance, and training; research and development of systems
for testing; proposal writing; writing for technical publications;
conference organizing, patents applications; acquiring research
funding, etc.
Photo by Frank Wojciechowski
Elmer Ledesma is a research staff member in the Department
of Mechanical and Aerospace Engineering.
Structure
Despite this amalgamation and diversity of functions, there
are established standards for these positions.
There is a difference between research staff and technical
staff, though the lines are often fuzzy. The focus of technical
staff is skill, while the focus of research staff is scholarship.
Technical staff master their craft, the equipment, and the
skills and enable the research of others. Research staff excel
scholastically: make discoveries, win patents, write papers,
and develop new research foci. Most research staff positions
require doctoral degrees. Though most professional research
staff are engaged in research programs directed by members
of the faculty, research staff in the upper
ranks may have the opportunity to lead their own research
programs. Senior research scientists may also become lecturers
for SEAS classes.
“Princeton has a commitment to being the best,”
Associate Dean of Faculty, Lin Ferrand *88 said. “That
commitment extends to the professional research and technical
staff.”
So that they may learn from one another’s methods and
research, research staff member Ihab Girgis organizes monthly
meetings of the PRTS in MAE.
Education
Although they are not faculty, the PRTS are valuable resources
for students.
At a recent meeting of the PRTS, Richard Miles explained that
since the SEAS is more career-oriented and reliant on practical
education than the rest of the University, the expertise of
nonacademics is especially important to SEAS students.
“The PRTSs significantly influence the undergraduate
experience,” Professor Miles said. “Students frequently
express appreciation for the research and technical staff.”
In MAE, technical staff members Glenn Northey, David Radcliffe,
and Mike Vocaturo are specifically designated to run the undergraduate
labs.
“The students know the abstract theories from class,”
Mr. Vocaturo said. “But experimental data differs from
abstract theories. So, we expose the students to both experiences.”
The undergraduate labs are brimming with tools and machinery.
Wind tunnels, water channels, gas turbine engines, hardness
testers, and even Legos—plenty of things students can
use to get their hands dirty.
“We like to make experiments very visible because seeing
is believing,” Mr. Vocaturo said.
These men are clearly proud of their students’ projects.
They are passionate about education and excel at it. Mr. Northey
received a Special Recognition Lifetime Achievement Excellence
in Teaching Award from the Engineering Council and the Princeton
University President’s Achievement Award.
Research
For most of the research and technical staff, however, it’s
the research itself that inspires them most.
For example, Seyed Allameh’s passion for his research
is obvious as he speaks about the new discoveries he and the
materials research group have made recently. Dr. Allameh normally
studies the fatigue and fracture properties of microelectromechanical
systems (MEMS). In December he and others in the thermostructural
materials group made a new discovery, which mere mention of
causes a wide, glowing grin to spread across Dr. Allameh’s
face.
“We stumbled upon something very new and exciting,”
Dr. Allameh said, clearly trying to contain his mirth.
Silicon MEMS are coated with a layer of platinum for protection.
Dr. Allameh and his colleagues were pleasantly surprised to
learn that simply by changing the angle at which to coat silicon
with platinum, the researchers can grow a peculiar phenomenon
they’ve termed “nanofins.” This growth vastly
increases the surface area of the MEMS, while reducing the
mass. This discovery could have grand implications for solar
panels or lightweight materials.
Dr. Ledesma, a research staff member in the combustion group,
has his Ph.D. in chemistry and did not expect to be part of
a mechanical and aerospace engineering department. Yet, assistant
professor Judy Wornat’s research drew him away from
his home in Australia. He now studies polycyclic aromatic
hydrocarbons (PAH), specifically the carcinogen catechol,
which is found in biomass combustion and in cigarettes.
“Everyone knows Princeton combustion,” Dr. Ledesma
said. “It’s very well-respected.”
Arnold Lettieri, a technical staff member III, first learned
of the Princeton Engineering Anomalies Research (PEAR) Lab
from a 1989 article in the New York Times Magazine. He was
fascinated by PEAR Lab’s study of human-machine interactions
and wanted to participate in that research.
Similar stories are told by many of the research and technical
staff members. Like the cogs and wheels in a clock, the PRTS
are always in action: synchronized and churning out information.
They are in the background and silent, except for that fervent
whir of discovery.
Looking
for more life in the Milky Way
Princeton vies to design NASA’s earth-finding
telescope
“Mankind will not remain
on Earth forever, but in its quest for light and space,
will at first timidly penetrate beyond the confines
of the atmosphere, and later will conquer for itself
all the space near the Sun.”
Konstantin Eduardovitch Tsiolkovsky, father of cosmonautics
(1857-1935)
This image is what a star would look
like viewed through the spergel pupil. The dark parts
on the side are the areas where researchers hope to
see planets.
|
Perhaps Tsiolkovsky wouldn’t have been surprised to
hear that man would walk on the moon in 1969. Yet, even he
may have been impressed to learn that less than a century
after his death, man would be looking past “all the
space near the sun” and focusing on space near other
suns far across the Milky Way galaxy.
The National Aeronautic and Space Administration (NASA), the
United States’ premier coalition of pioneers, has begun
a new project that could answer the very question that probably
inspired the first astronomers to stare wonderingly into the
stars: Is there other life in the universe? NASA’s new
initiative, the Terrestrial Planet Finder (TPF), might find
out.
The TPF will be a space-based telescopic camera. NASA’s
intent is that the TPF will find 150 planets—and hopefully
one that exhibits characteristics that indicate life could
exist. By examining the diffraction of light in the picture
of a planet, biologists and atmospheric chemists can detect
the presence of bioindicators, compounds that are essential
to life. Recent studies have disclosed that some bioindicators
can be viewed in visible light—not just infrared—as
was once thought.
In 2000 NASA launched a study, challenging four industry teams
to study possible design concepts for the best TPF. After
narrowing the scope to the most feasible designs, NASA selected
six teams from industry and academia to delve further into
these options. A group from Princeton, working alongside Ball
Aerospace and Technologies Corp., is one of these six.
Professor Jeremy Kasdin ’85 from the Department of Mechanical
and Aerospace Engineering (MAE) is principal investigator
of the Princeton–Ball study. He is backed up by an all-star
team, including Professors Michael Littman of MAE, Robert
Vanderbei of operations research and financial engineering,
and David Spergel ’82 and Edwin Turner of astrophysics.
Also on the team are research staff members Daniel Mumm and
Pinchas Gurfil from MAE, Michael Carr from astrophysics, and
Sara Seager from the Institute for Advanced Study and graduate
students Amir Give’On, Russ Arrell, and Eric Ford. A
big group, yes. But each one will be needed if they’re
to finish all the work ahead of them between now and October.
“We want to show at the end of this six months that
we’re moving in the right direction,” Professor
Kasdin said. “We want NASA to believe that it would
be foolish to stop, and that there’s something going
on here worth continuing.”
That “something” is called the Spergel Gaussian
Pupil Optical Coronagraph, named after Professor Spergel.
The concept, in a nutshell, is a large telescopic camera that
views planetary systems in the visible light, and via a special
aperture shaped like a cat’s eye, and suppresses the
glow of the side lobes of a star’s Aery pattern so that
the glow of a planet can be seen.
So, what does that mean?
The theory of ray optics states that when light passes through
an object with a greater density—like a lens—the
light slows down, which causes it to refract. The light is
thus viewed in the “image plane,” looking precisely
as it did on the opposite side of the lens, only upside down.
So theoretically a star viewed at such an extraordinary distance
would appear in the image plane as a simple point of light.
“In fact, this is a great idealization of how lenses
operate,” Professor Littman said.
No lens can identically reproduce an image. Errors in the
lens create distortions in the viewable images. Because of
the fixed size of the lens, a point of light ends up looking
like a bull’s-eye—bright at the center, and getting
progressively dimmer toward the outside—a phenomenon
called the Aery pattern. A star is about 10 billion times
brighter than a planet. Even though the outside rings of the
Aery pattern aren’t quite as bright as the center point,
they are still brighter than a dim little planet nearby. If
the TPF is ever going to find the little planets, it’s
got to blot out the star’s overwhelming glow.
When the team began meeting in 2000, they had this daunting
challenge cavalierly lounging in front of them. The Spergel
pupil was not even a glimmer in their minds.
“The problem we’re attacking is one that no one’s
ever worried about before,” Professor Kasdin said. “Most
of the time, people are trying to get a better picture of
the star. We don’t really care at all what the star
looks like.
“We would meet and just brainstorm, asking ‘How
are we going to solve this problem?’” Professor
Kasdin said. “We’d all decided that the leading
concept at the time was just excruciatingly hard, and probably
not a practical way of doing it. So we just started tossing
ideas around.”
“It was fun because we were brainstorming all kinds
of things,” Professor Littman said. “Usually you
kick around ideas with people who are in fields similar to
what you’re doing. Here, our group was really interdisciplinary.”
“It’s very hard for people to think outside the
box of their own field,” Professor Kasdin said, “because
they have all these biases and prejudices. But someone in
another field could say, ‘Why don’t you try this
completely crazy, ridiculous thing?’”
“And after you say, ‘You idiot!’”
Professor Littman said, “You say, ‘Well, you know.
That’s not that bad an idea.’”
It was in one of these meetings that Professor Spergel came
up with something the engineers thought was a crazy, ridiculous,
not-that-bad idea.
Professors Kasdin and Spergel were experimenting with a concept
of optics. By grouping several telescopes together, they act
like one larger telescope. Different grouping arrangements
create different light patterns in the image plane. Professor
Spergel took this idea a step further.
“It was Dave’s inspiration,” Professor Kasdin
said. “He looked at it and said, ‘Well, if I’m
just going to take a bunch of satellites and put them together
in a pattern, it’s no different if I just make a pupil
that shape. Mathematically it’s the exact same thing,
and much easier to do.’ It was just one of those light-bulb
moments.”
The pupil is an elliptical shape, wide at the center, and
pointed at the ends, like a cat’s eye. This is achieved
simply by placing an opaque mask over a regular round telescopic
mirror.
Optic principles dictate that this oddly shaped aperture will
allow two fans of light to shine above and below the point,
while suppressing the side lobes where planets are likely
to be hiding.
NASA was impressed. The concept is simple and therefore comparatively
inexpensive—and it seemed to work. Unfortunately, optic
imperfections cause more problems.
The special pupil suppresses starlight so planets can be seen,
but the mirror errors are still there. Pictures of the planets
will be distorted, making it impossible to detect bioindicators.
The Princeton group had to ask, “How do we counteract
these errors?”
As embattled optic scientists have known for years, there
are two basic types of errors that a mirror can have: phase
errors and amplitude errors.
Phase errors are caused by bumps, dimples, ridges, and trenches
in the surface of the mirror. These errors cause the waves
of light reflected off the mirror to be out of sync. Each
peak of each wave should be in line with the one below it,
but phase error scoots them out of alignment.
Amplitude error is caused by varying reflectivity in the mirror
surface. In that case, the shape of the waves may not appear
accurately.
The team’s plan to counteract these errors is based
upon methods used on ground-based telescopes to null distortions
caused by the atmosphere. Two flexible, deformable mirrors
—or DMs—one for phase and one for amplitude, will
re-refract the image, adjusting as necessary to correct the
flaws in the rigid telescopic mirror. Thus, the final picture
is as close to perfect as possible. Except …
“Let’s assume we have a DM that does what we want
it to do,” said Professor Kasdin. “How do I figure
out, from the distorted light in the image plane how to move
the DM? All we see is a bunch of light scattered across the
camera. From that we’ve got to figure out what that
image implies about errors in the optics so that we know how
to move the DM.”
This last problem is the group’s focus at the moment.
Mr. Give’On’s work has been making progress in
this direction, but it is still in the early stages. Nonetheless,
Professor Kasdin is confident that the team will solve all
their difficulties and come out on top.
“We believe we can get this to work, and get it to work
in space,” he said. “There’s a hundred ways
to do it. We just think we’ve come up with the coolest
way.”
NASA intends to make a final decision on the design by 2006.
Photo by Frank Wojciechowski
Professors Jeremy Kasdin, Robert Vanderbei, and Michael Littman
are part of a research effort that may help discover answers
to a question asked throughout the ages: Are we alone in the
universe?
Optimizing
simulator breaks data sets into ‘little pieces’
for manageability
At heart, Warren Powell ’77 is a true academic. He admits
that he likes “elegant mathematics” and loves
writing papers. Yet, he has more intimate ties with industry
than many of his colleagues in academia.
Professor Powell runs CASTLE Lab in the Department of Operations
Research and Financial Engineering (ORFE). CASTLE, which stands
for Computational and Stochastic Transportation and Logistics
Engineering, maintains corporate partnerships with six freight
transportation organizations.
The flow of Norfolk Southern Railroad’s
locomotives is a complex web across the East.
The matching of individual locomotives
with individual trains is represented by this mesh.
This graph shows a snapshot of flows
at Burlington Northern and Santa Fe Railroad.
|
“It’s a mouthful,” Professor Powell said.
“The theme of research in this lab is learning how to
model the organization and flow of decisions and information.”
CASTLE Lab studies the operations of freight transportation
organizations. To help their human workers make better decisions,
these freight organizations use tools that simulate the physical
process of moving cargo from place to place. Sometimes, CASTLE’s
ultimate goal is to supply the corporate partners with new
decision-making software, which uses these better tools.
CASTLE Lab was officially established in 1992, although Professor
Powell began his first corporate partnership with Yellow Freight
System (now Yellow Transportation), a trucking company, in
1989.
Corporate partners not only fund research, but also provide
feedback about the findings. The lab develops technologies
that are both mathematically sensible and feasible from a
business standpoint. The lab work is generalized, so many
partners benefit from the same research.
Research
The focus of the CASTLE Lab research is what Professor Powell
has termed “optimizing simulators.”
Tools for simulating and decision making generally fall into
two categories: simulation systems or optimization systems.
Simulators implement a lengthy set of rules that mimic the
decisions made by humans running a system.
Optimizers use mathematical algorithms to quickly identify
the best solution. Simulators can capture a high level of
detail, but exercise a low level of intelligence. Optimizers
have a high level of intelligence, but require many simplifying
assumptions.
“The Air Force has been talking to various academics
over the past decade to bring in smarter tools,” Professor
Powell said. “When I became involved in this discussion,
I came in and said, ‘Gee. Simulation vs. optimization.
Why are we asking this question?’
“I have the technology that simulates, but with a level
of intelligence that can actually compete with an optimizer.”
This system can actually perform with up to 99.99 percent
of the intelligence of an optimizer. So how does Professor
Powell do it?
“Rather than optimizing the whole thing, I optimize
little pieces,” he said.
An optimization system looks at an entire set of data all
at once, and out of the millions of possible solutions it
chooses the “best” one to minimize cost and maximize
profit. However, to do so, it must assume that all the data
are perfect and disregard the flow of information over time.
Unfortunately for optimization systems, data are never perfect
and actions in the present do affect actions in the future.
The optimizer simulator breaks down the entire data set into
little pieces in time and space.
For example, Norfolk Southern Railroad has many stations around
the country operating all day. One piece of this entire data
set is one Norfolk Southern station at a particular point
in time. The simulator enters all the information, such as
the number of trains, the types of locomotives, the destinations,
etc. Although there are many details, the optimizer can handle
them, because the data set is small. The optimizer must decide
how many locomotives of each type to attach to each train,
en route to each destination. Of course, this decision changes
the distribution of locomotives from station to station, later
in time.
So, the simulator runs repeatedly, using feedback mechanisms
to ensure that all the pieces of the puzzle act as a cohesive
unit. Thus, Norfolk Southern gets the best of both the simulator
and the optimizer worlds.
As with most everything, this is easier said than done. Sometimes
decisions must be made when one is not privy to all the factors
influencing the outcome of the decision. The two main problems
are the information that one will never know—incomplete
information—and the information that one will eventually
know, but too late —uncertainty.
“We had to come up with a way to solve this just to
survive. Otherwise, the projects would fail,” Professor
Powell said.
These issues required a great deal of time and thought. Professor
Powell enlisted the help of Arun Marar *02, who was a member
of the technical staff and doing a dissertation on incomplete
information. Huseyin Topaloglu *00 focused his doctoral research
on optimizing under uncertainties.
Dr. Marar’s dissertation explains that incomplete information
is never directly visible, but can be inferred by studying
the history of human decisions and detecting patterns in their
action.
“How do we handle something that we just don’t
have data on?” Professor Powell asked. “What we
have is what humans did in the past. We bring these patterns
into the model.”
Professor Powell used the following example. A freight company
may normally prefer to match up locomotive type 1 with train
type A. However, if train A must travel on course alpha, then
locomotive type 1 cannot be used because that locomotive can’t
handle the tight turns on the chosen course. Neither a standard
optimizer nor simulator could process details such as the
geography of each individual course. A human operator can,
however, and easily will make an adjustment to the normal
rules. CASTLE’s operations researchers can detect these
patterns in the human behavior and incorporate them into the
stochastic modeling software they create.
Dr. Topaloglu’s work produced algorithms that quickly
solve uncertainty problems, by weighing the probabilities
of different outcomes. These efforts made it possible for
CASTLE Lab to clear some daunting hurdles.
“By doing it here in the university, we’re able
to get an elegant theory,” Professor Powell said. “The
academic environment offers a different culture than industry.
It’s more focused on creativity.”
Conversely, the corporate partnerships supply CASTLE Lab with
every professor’s essential tool: a teaching aid.
Education
CASTLE’s software is valuable to students as well as
CEOs.
In ORF411: Operations and Information Engineering, students
play the Orange Juice Game, which draws on the same software
library used by the railroads. The students simulate the running
of an orange juice company, with the goal of running the company
as efficiently as possible while serving the most people at
the lowest cost.
Students in ORF411 called the game “meaningful,”
“stimulating and tangible,” “a critical
tool,” and “the highlight of the semester.”
“I got a glimpse of what it would be like to manage
the operations of a real company,” Scott Dias ’02
said.
Adrienne Clark ’02 said, “I gained a much greater
appreciation for the real-world application of my course work.”
Students value theoretical education more highly when they
can see theories applied. CASTLE Lab’s work helps Professor
Powell illustrate certain business realities that most textbooks
do not cover.
If CASTLE Lab workers simply ignored the practical problems
of industry and used data sets and assumed perfect data, then
questions of incomplete data and uncertainty would never be
solved, and corporations wouldn’t be able to apply elegant
theories to everyday use.
The optimizing simulator concept is currently running at Yellow
Transportation, Norfolk Southern Railroad, and Burlington
Northern and Santa Fe Railroad. Other partners are Air Products
and Chemicals, Triple Crown Services, and the Air Mobility
Command. The Air Force Office of Scientific Research is a
substantial supporter of CASTLE Lab’s research.
The combination of university time and minds and industrial
data and feedback has been essential to CASTLE Lab’s
success.
“The result is not just mathematical elegance,”
Professor Powell said. “We actually end up with tools
that work.”
This image represents the flow of Yellow Transportation’s
drivers across the country, color-coded by domicile.
CEE
researchers working to fit pieces of global warming puzzle
together
In modern times, children are taught at an early age about
global warming. They hear ghastly tales about the future after
climate change goes too far.
They see delicate warm-blooded creatures dying of the heat,
and armies of hardy, metal-encased insect knights taking over
the planet. They see the polar ice caps melting, flooding
the coastlines, and the Statue of Liberty buried up to her
neck in dim, brackish waters.
And further inland, gone are the meadows rife with flowers.
In their place sprawl dusty deserts, littered with the bones
of animals that starved as their food supply slowly dwindled.
The predictions of our future after major climatic change
are grim.
But is climate change really happening?
Across the globe, large-scale collaborative efforts are being
made to answer this question. Scientists specializing in many
disciplines are adding their input, hoping it will add up
to some answers.
The World Climate Research Programme’s Global Energy
and Water Experiment (GEWEX) is enlisting the help of scientists
worldwide, including a group from Princeton, who say that
it may be too early to tell whether the climate is changing
or not.
Civil and Environmental Engineering Professor Eric Wood and
researchers at the Program in Environmental Engineering and
Water Resources are numbered among GEWEX’s forces.
Their individual focus is on land surface-atmosphere interactions,
and they aim to determine whether the terrestrial water and
terrestrial energy cycles are changing.
The National Aeronautics and Space Administration (NASA) stated
that the most “significant manifestation” of climate
change would be an increase in the rate of the water cycle.
Since evaporation is both a major transfer of water and energy,
the two cycles are closely linked and studied together.
If scientists can develop an accurate model of these systems
that can be applied to any region on Earth, this will help
show them whether or not these cycles—and therefore,
the climate—are indeed changing.
Before perfecting such a model they must first develop more
sophisticated data collection tools and determine where and
how to perform testing.
GEWEX is coordinating efforts by researchers who are collecting
data on precipitation, runoff, evaporation, and soil moisture.
“The GEWEX activities are basically used to try and
understand these systems, through observation and modeling,”
Professor Wood said, “and then ask questions such as
‘Is the hydrological cycle changing or not?’”
First, data collection tools need to be improved. NASA satellites
are currently collecting data, but the enormous footprint
of a satellite makes for coarse information. Small-scale research
is necessary to verify the accuracy of the satellite.
“The remote-sensing can work over large scales, which
would help, because you can’t measure everywhere by
building towers and digging holes all over the place,”
Professor Wood said. “Can we observe enough through
space observation alone? That would be good, because we would
have consistent measurements. We wouldn’t have to actually
go to Afghanistan to measure the rainfall there.”
Early in May NASA launched AQUA, a satellite designed to remotely
collect data on soil moisture. The satellite infers soil moisture
from data on the radiation emitted from the land surface.
Professor Wood’s research group is doing fieldwork in
Iowa, collecting soil moisture data, comparing it to the satellite
data, and verifying the accuracy of the satellite measurements.
“It’s a logistic question,” he said. “Do
we have the instruments? Are they sufficiently accurate?”
Let’s assume that all the tools work perfectly. There
are still the questions of where to test and how much data
to collect.
Confident predictions take time. The more erratic something
is, the longer it takes to detect a trend.
As a hypothetical example, picture your favorite village in
Ghana. Village records from the past 100 years indicate that
the local precipitation has ranged widely, vary ing by as
much as 40 percent from year to year. Then, for five consecutive
years, the precipitation is substantially higher than the
mean of the past 50 years. Does this mean that the village
climate is getting wetter? That the hydrological cycle is
accelerating? Maybe. Maybe not. It’s simply too early
to tell because a village with such unpredictable rainfall
may just be exhibiting more of its natural variability.
“If you want higher confidence, you have to wait longer
periods of time than you would if you relaxed a little,”
Professor Wood said. “I could say, ‘I might err
a little bit. I’ll say there’s climate change
happening, when it may not really be happening. Maybe we would
rather err on the side of being careful.’”
By measuring the natural variability of three major components
of the hydrological cycle—precipitation, evaporation,
and runoff—Professor Wood’s group has estimated
the number of years required to detect substantial changes
in the trends of these components (see graphs).
Each climatic element has its own individual variability.
So even if the precipitation of a region is relatively stable,
the runoff could be widely erratic, raising the study time
higher and higher. North America’s most variable component
is runoff, due to the Mississippi River, a force that dominates
the entire continent—hydrologically speaking. Professor
Wood estimates it would take 73 years of data to confidently
detect a trend in the runoff of North America. Still, this
is an easy wait compared to the 173 years it will take to
gather enough information on African precipitation.With such
fluctuation in smaller regions, how can hydrologists accurately
determine climate change across the globe? Most research focuses
on the most exciting, dynamic areas of the world, ones with
mighty rivers and waterfalls. Should it?
“If you look at the Western Hemisphere’s big
study regions, you can ask yourself ‘Are these river
basins representative of what’s happening on the continent?’”
Professor Wood said. “And it turns out that the trends
for North America [as a whole] are quite different. The sites
aren’t that representative.”
Professor Wood and his group have used an optimization system
to determine which sites to study so that they can paint a
more accurate, comprehensive picture of the whole. The optimizer
chooses a sample set that accurately represents the greater
region, the number of sites equaling five percent of all possible
test sites.
Then, field testing in these areas is needed. The experimental
data is then worked into the models.
“These processes are very complicated,” Professor
Wood said, “and we just don’t understand all the
physics yet.”
Certainly, many mysteries remain for climate change scientists.
How do we perfect the technology? Are the models telling us
the truth? What sort of sobering information will we learn
when it’s all said and done? All Professor Wood’s
work will help create better tools so that we can get better
answers. Once some of these mysteries are solved, we’ll
have ways to combat the climate change specter that haunts
our dreams of the future.
Starting
from scratch
Professor Ruby Lee rethinks computer design
Ruby Lee wants a new computer. She’s not thinking about
a new machine for her office in the electrical engineering
department. She is entirely rethinking the way computers are
designed.
Professor Lee, the Forrest Hamrick Professor of Engineering,
has an ambitious research project to develop from scratch
the core elements of a computer so they are far better at
dealing with security, privacy, and new media: integrated
sound, pictures, video, and text—all things Professor
Lee believes will be crucial in the coming decades.
“I am interested in designing the computer processor
architecture for the 21st century,” Professor Lee said.
“What would that processor look like if we could design
it from scratch?”
It may sound like a tall order, but in some ways she has done
it before. As a chief computer architect at Hewlett-Packard,
she was a key figure in a revolution in computer architecture
that swept through the industry in the 1980s. With other pioneers,
she advocated the use of a vastly simplified set of core instructions
that computers use in carrying out all the complex things
they can be programmed to do. This simplified system is now
at the heart of tens of billions of dollars worth of computer
systems sold each year.
“She’s really been one of the top instruction-set
architects in the world,” said Joel Birnbaum, the former
senior vice president at Hewlett-Packard who hired Professor
Lee in 1981. Before she left the company to come to Princeton
in 1998, Professor Lee was leading a team that collaborated
with chip maker Intel Corp. to design a new architecture,
which was recently released in the new “Itanium”
microprocessor chips and which Mr. Birnbaum said will remake
the industry once again.
“Ruby has been there for two major revolutions in architecture,”
he said. “That’s quite unusual for a computer
architect.”
But that’s not to mention a third innovation Professor
Lee brought to the industry. She also led the way in creating
a set of multimedia instructions that build directly into
the core native language of the computer the ability to handle
multimedia of all types—images, voice, and animation.
Intel later adopted the idea, which consumers saw when the
company began advertising that its chips had multimedia-accelerating
“MMX” technology. Now essentially all computers
use the technique, allowing multimedia to be an everyday part
of computing.
Princeton Professor of Computer Science Kai Li called Professor
Lee’s work in this area a “seminal contribution”
that reflected her ability to predict very early on how important
graphics and multimedia would become.
It also is impressive, said Princeton Electrical Engineer
Sun-Yuan Kung, that Professor Lee’s solution was compatible
with all the existing technology. “That was really a
smart move and was a great contribution,” he said.
Clean slate
Having reached this point, however, Professor Lee is ready
to start over.
Moving from industry to academia allowed her to “start
from a clean slate,” she said. “In industry you
can rarely start from a clean slate,” she said, noting
that successfully entrenched products must always build on
the previous versions for compatibility reasons. Now that
she is here, she said, “I find that freedom exhilarating.”
In her reexamination, one of the first things she found lacking
is the way computers handle security. She noted that in the
early days of computers, security amounted to putting the
machines in a locked room. Now that computers are connected
all the time to networks, including the global network of
the Internet, security is a much more difficult and crucial
issue.
New security tools
Current security tools, such as virus-checkers, computer “firewalls,”
and encryption techniques, are a patchwork that can be difficult
to use and are far from comprehensive. Professor Lee wants
to make security part of the basic architecture of the computer,
just as she did with multimedia.
“We need security that is implicit and invisible,”
she said, “not security that is a negative impact on
our quality of life—like having to line up for a long
time at the airport to go through checks.
Real security is built in, not added on as an afterthought.”
On the multimedia front, Professor Lee is developing what
she calls the “canonical, minimalist instruction-set
architecture for multimedia processing that is fully general
purpose, high-performance, low-cost, and low-power, so it
can be implemented in the smallest information appliance,
like a Dick Tracy watch.”
In both of these pursuits, Professor Lee’s approach
is to develop technology that does more than just run fast.
“I think about how you can make a processor more effective,
rather than just more efficient,” she said.
The goals may sound lofty, but it would be a mistake to assume
that Professor Lee is no longer interested in whether her
ideas will win in the marketplace.
“I can’t get away completely from the idea of
my research being important to the future of the computing
industry,” she said.
In her latest innovation, she discovered a way to reduce by
100-fold the number of operations a computer must execute
to scramble a series of bits, the smallest units of information
in a computer.
Current computers are best at processing word-sized chunks
of information, where each word is typically 64 bits. This
design, however, is inefficient at manipulating bits within
a word, which can be valuable for encrypting data such as
credit card information sent over the Internet. A typical
computer might have to execute hundreds of instructions to
arrive at a specific arrangement, or permutation, of a 64-bit
unit.
Professor Lee and her students recently published papers showing
how computers of the future could be modified to perform 64-bit
permutations in, at most, six instructions, while keeping
all the benefits of word-oriented processors. Since submitting
that paper, she has further reduced the process to just one
step, she said.
“I would like to start a conversation, first at Princeton,
then nationally and internationally, to answer the very basic
questions of who should be allowed to access what type of
information and when. Then the technologists can respond with
what can or cannot be achieved, or what else can be achieved.”
Back to academics
After earning her 1980 Ph.D. from Stanford University, Professor
Lee stayed on for 15 months as an assistant professor until
Mr. Birnbaum recruited her to Hewlett-Packard. During her
years in industry, Professor Lee kept in touch with her academic
counterparts and maintained a consulting professor relationship
with Stanford.
Nonetheless, Professor Lee encountered some surprises when
she returned full-time to academia. One was learning about
the extent of the inequities women still face in rising through
the science and engineering fields. Shortly after she arrived,
former president Harold Shapiro *64 invited her to a conference
on the subject at the Massachusetts Institute of Technology.
Now she is a member of a task force on gender equity in the
natural sciences and engineering established by President
Shirley Tilghman.
“I thought there wasn’t much of a problem in Silicon
Valley—at least I never felt much of an impact,”
she said. “But I have seen enough data to convince me
that there is a real problem with the biases women face, and
hence in attracting and keeping women in academic science
and engineering fields. This is a problem that seems incongruous
with the 21st century—we must find creative and equitable
solutions.”
Professor Lee is excited about another aspect of being at
the University: the students.
“The greatest challenge in teaching is to give these
very bright students the most important concepts we have learned
over the years, and train them to think out of the box, because
we don’t want the next generation of computer architects
designing more of the same,” she said.
She also has enjoyed recruiting top graduate and undergraduate
research students to her quest for new and better ways of
designing computers.
“The students are fearless because they don’t
know what can’t be done,” she said.
Perhaps one of them will build her a new computer.
[ contents
] [
previous story ] [
next story ]
[ top
of page ]
|
